Functional Importance of GGXG Sequence Motifs in Putative

Jul 13, 2009 - Adam Dobrowolski and Juke S. Lolkema*. Molecular Microbiology, Groningen Biomolecular Sciences and Biotechnology Institute,. University...
0 downloads 0 Views 2MB Size
7448

Biochemistry 2009, 48, 7448–7456 DOI: 10.1021/bi9004914

Functional Importance of GGXG Sequence Motifs in Putative Reentrant Loops of 2HCT and ESS Transport Proteins† Adam Dobrowolski and Juke S. Lolkema* Molecular Microbiology, Groningen Biomolecular Sciences and Biotechnology Institute, University of Groningen, Haren, The Netherlands Received March 23, 2009; Revised Manuscript Received July 10, 2009 ABSTRACT: The 2HCT and ESS families are two families of secondary transporters. Members of the two families are unrelated in amino acid sequence but share similar hydropathy profiles, which suggest a similar folding of the proteins in membranes. Structural models show two homologous domains containing five transmembrane segments (TMSs) each, with a reentrant or pore loop between the fourth and fifth TMSs in each domain. Here we show that GGXG sequence motifs present in the putative reentrant loops are important for the activity of the transporters. Mutation of the conserved Gly residues to Cys in the motifs of the Naþcitrate transporter CitS in the 2HCT family and the Naþ-glutamate transporter GltS in the ESS family resulted in strongly reduced transport activity. Similarly, mutation of the variable residue “X” to Cys in the N-terminal half of GltS essentially inactivated the transporter. The corresponding mutations in the N- and C-terminal halves of CitS reduced transport activity to 60 and 25% of that of the wild type, respectively. Residual activity of any of the mutants could be further reduced by treatment with the membrane permeable thiol reagent N-ethylmaleimide (NEM). The X to Cys mutation (S405C) in the cytoplasmic loop in the C-terminal half of CitS rendered the protein sensitive to the bulky, membrane impermeable thiol reagent 4-acetamido-40 -maleimidylstilbene-2,20 -disulfonic acid (AmdiS) added at the periplasmic side of the membrane, providing further evidence that this part of the loop is positioned between the transmembrane segments. The putative reentrant loop in the C-terminal half of the ESS family does not contain the GGXG motif, but a conserved stretch rich in Gly residues. Cysteine-scanning mutagenesis of a stretch of 18 residues in the GltS protein revealed two residues important for function. Mutant N356C was completely inactivated by treatment with NEM, and mutant P351C appeared to be the counterpart of mutant S405C of CitS; the mutant was inactivated by AmdiS added at the periplasmic side of the membrane. The data support, in general, the structural and mechanistic similarity between the ESS and 2HCT transporters and, more particularly, the two-domain structure of the transporters and the presence and functional importance of the reentrant loops present in each domain. It is proposed that the GGXG motifs are at the vertex of the reentrant loops.

The 2HCT [2-hydroxycarboxylate transporter, TC 2.A.24 (1)] and ESS (glutamate:Naþ symporter, TC 2.A.27) families represent families of ion-driven transporter proteins that are exclusively found in the bacterial domain. Members of the two families do not share any significant amino acid sequence similarity, but the hydropathy profiles of the sequences are very similar. For this reason, the two families are in the same structural class (ST[3]) in the MemGen classification system that we have introduced to identify membrane proteins sharing the same fold (2-5). The MemGen classification system groups membrane proteins in structural classes based on hydropathy profile analysis. The hydropathy profile of the amino acid sequence of the membrane protein is taken to be characteristic for the folding of the protein in the membrane. Recently, strong support for the MemGen classification was obtained by the similar organization of the core in the high-resolution structures of members of the NSS [neurotransmitter sodium symporter, TC 2A.22 (6)], SSS [sodium †

This work was supported by grants from The Netherlands Organization for Scientific Research (NWO-ALW). *To whom correspondence should be addressed: Molecular Microbiology, Biological Centre, Kerklaan 30, 9751 NN Haren, The Netherlands. Phone: þ31 50 363 2155. Fax: þ31 50 363 2154. E-mail: j.s. [email protected]. pubs.acs.org/Biochemistry

Published on Web 07/13/2009

solute symporter, TC 2A.21 (7)], and NCS1 [nucleobase cation symport 1, TC 2A.39 (8)] transporter families. While the members of the NSS, SSS, and NCS1 families do not share sequence similarity, the families are all found in the same structural class (ST[2]) in the MemGen classification (3, 9). Experimental support for the same fold of the proteins in the 2HCT and ESS families was obtained by demonstrating a similar membrane topology for two transporters from the two families (10). The well-established membrane topology model of the 2HCT family, mostly based on studies of Naþ-citrate transporter CitS of Klebsiella pneumoniae (reviewed in ref 11), was used to predict the membrane topology of Naþ-glutamate transporter GltS of Escherichia coli (12, 13), a member of the ESS family. The model was verified by accessibility studies of cysteine residues introduced into the GltS protein (10). Though membrane topology of a protein represents a low structural resolution, the result is not trivial and does validate the MemGen classification because of specific structural features of these proteins. Secondary structure predictors like TMHMM (14) predict different models for both CitS and GltS that were inconsistent with the experimental data (10). The structural model of the transporters in the ESS and 2HCT families, and in all 33 families of structural class ST[3] in the r 2009 American Chemical Society

Article

Biochemistry, Vol. 48, No. 31, 2009

7449

FIGURE 1: (A) Structural model for the transporters of the 2HCT and ESS families. Two homologous domains containing five TMSs each with an inverted topology in the membrane are indicated in dashed boxes. Each domain contains a pore loop structure entering the membrane-embedded part of the protein from the periplasmic and cytoplasmic sides of the membrane, respectively (Vb and Xa). Members of the 2HCT family have an additional TMS at the N-terminus that is not present in members of the ESS family. (B) Sequence logos of regions Vb and Xa showing the GGXG sequence motifs in the 2HCT (left) and ESS (right) families. Position numbers correspond to the residue numbers in the CitS (2HCT) and GltS (ESS) sequences. Residues mutated into cysteines are indicated in the red dotted boxes. The logos were generated using WebLogo, version 2.8.1 (http://www.bio.cam.ac.uk/cgi-bin/seqlogo/logo.cgi).

MemGen classification, consists of two domains each containing five transmembrane segments (Figure 1A). The total number of transmembrane segments is variable between different families. For instance, the 2HCT proteins have an additional segment at the N-terminus which is missing in the ESS proteins. In the model, the two domains of five TMSs each share a similar fold but have opposite orientations in the membrane (11, 15, 16), a structural motif seen more frequently in membrane proteins (inverted topology) (6-8, 17-20). The loops between the fourth and fifth transmembrane segments in each domain are believed to form so-called pore loops or reentrant loops, which fold back between the transmembrane segments from opposite sides of the membrane [trans-reentrant loops (15, 16)]. The reentrant loops in the N- and C-terminal domains are believed to be in close vicinity in the three-dimensional structure and to form the translocation pathway for substrate and co-ions. The putative reentrant loop regions, termed Vb in the Nterminal domain and Xa in the C-terminal domain, are well conserved within the families throughout structural class ST[3] and contain a remarkably high fraction of residues with small side chains like Gly, Ala, and Ser (11). In almost all families of ST[3], small stretches of highly conserved residues are found in putative reentrant loop regions Vb and Xa. Sequence analysis of the 138 members of the ESS family and the 74 members of the 2HCT family showed that in the Vb regions of both families and the Xa region of the 2HCT family these conserved stretches

contain a GGXG sequence motif in which X is a less conserved residue (see Figure 1B). The same motif cannot be found in the Xa region of the ESS family, but also here, three conserved Gly residues are found in a stretch of eight residues. Here, we present a mutational study of the GGXG sequence motifs found in the putative reentrant loops of the CitS and GltS proteins in determining the relevance of the motifs for the transport function of the proteins. The functional relevance of the Xa region in GltS in which the motif is not found is addressed by cysteine-scanning mutagenesis. It follows that the motifs play an important role in the transport mechanism catalyzed by both transporters, and further evidence for the existence of the reentrant loops is obtained. The corresponding properties of CitS and GltS further support a similar core structure and mechanism for the two transport proteins. EXPERIMENTAL PROCEDURES Bacterial Strains, Growth Conditions, and GltS and CitS Constructs. E. coli strain DH5R was routinely grown in LuriaBertani broth (LB) medium at 37 °C while being continuously shaken at 150 rpm. Ampicillin was used at a final concentration of 50 μg/mL. The GltS and CitS proteins were expressed in E. coli DH5R cells harboring plasmid pBAD24 (Invitrogen) derivatives encoding the wild type or cysteine mutants of GltS (10) and CitS (23) extended with six additional histidine residues at the N-terminus (His tag). In the case of CitS variants, a sequence

7450

Biochemistry, Vol. 48, No. 31, 2009

encoding an enterokinase cleavage site was present between the His tag and the citS gene. Expression of genes cloned in pBAD24 is under the control of the arabinose promoter. Production of the GltS and CitS proteins was induced by addition of 0.01% arabinose when the optical density of the culture measured at 660 nm (OD660) reached a value of 0.6. The cysteine mutants of GltS and CitS were constructed by PCR using the QuickChange site-directed mutagenesis kit (Stratgene, La Jolla, CA). All mutants were sequenced to confirm the presence of the desired mutations (ServiceXS, Leiden, The Netherlands). Transport Assays in RSO Membranes. E. coli DH5R cells expressing CitS or GltS variants were harvested from a 1 L culture by centrifugation at 10000g for 10 min at 4 °C. Right-sideout (RSO) membrane vesicles were prepared by the osmotic lysis procedure as described previously (21). RSO membranes were resuspended in 50 mM KPi (pH 7), rapidly frozen, and stored in liquid nitrogen. The membrane protein concentration was determined by the DC protein assay kit (Bio-Rad Laboratories, Hercules, CA). Uptake by RSO membranes was assessed by the rapid filtration method. The membranes were energized using the potassium ascorbate/phenazine methosulfate (PMS) electron donor system (22). Membranes were diluted to a final concentration of 0.5 mg/mL in 50 mM KPi (pH 6.0) containing 70 mM Naþ, in a total volume of 100 μL at 30 °C. Under a constant flow of watersaturated air, and while the mixture was being stirred magnetically, 10 mM potassium ascorbate and 100 μM PMS (final concentrations) were added, and the proton motive force was allowed to develop for 2 min. Then, L-[14C]glutamate or [1,5-14C]citrate was added at a final concentration of 1.9 or 4.4 μM, respectively. Uptake was stopped by the addition of 2 mL of ice-cold 0.1 M LiCl, followed by immediate filtration over cellulose nitrate filters (0.45 μm, pore size). The filters were washed once with 2 mL of a 0.1 M LiCl solution and assayed for radioactivity. The background was estimated by adding the radiolabeled substrate to the vesicle suspension after the addition of 2 mL of ice-cold LiCl, immediately followed by filtering. Partial Purification of GltS and CitS Derivatives by Ni2þ-NTA Affinity Chromatography. E. coli DH5R cells expressing CitS or GltS variants were harvested from a 200 mL culture by centrifugation at 10000g for 10 min at 4 °C. Cells were washed with 50 mM KPi buffer (pH 7) and resuspended in 2 mL of the same buffer and, subsequently, broken with a Soniprep 150 sonicator operated at an amplitude of 8 μm by nine cycles consisting of 15 s ON and 45 s OFF. Cell debris and unbroken cells were removed by centrifugation at 9000 rpm for 5 min. Membranes were collected by ultracentrifugation for 25 min at 80000 rpm and 4 °C in a Beckman TLA 100.4 rotor and washed once with 50 mM KPi (pH 7.0). His-tagged GltS and CitS derivatives were partially purified from the cytoplasmic membranes or RSO membranes prepared as described above using Ni2þ-NTA affinity chromatography as follows. Membranes (4 mg/mL) were solubilized in 50 mM KPi (pH 8), 400 mM NaCl, 20% glycerol, and 1% Triton X-100 followed by incubation for 30 min at 4 °C under continuous shaking. Undissolved material was removed by ultracentrifugation at 80000 rpm for 25 min at 4 °C. The supernatant was mixed with Ni2þ-NTA resin (50 μL bed volume per 5 mg of protein), equilibrated in 50 mM potassium phosphate (pH 8.0), 600 mM NaCl, 10% glycerol, 0.1% Triton X-100, and 10 mM imidazole, and incubated overnight at 4 °C under continuous shaking. Subsequently, the column material was pelleted by pulse

Dobrowolski and Lolkema centrifugation, and the supernatant was removed. The resin was washed with 10 volumes of equilibration buffer containing 300 mM NaCl and 40 mM imidazole. The protein was eluted with one-half of a bed volume of the washing buffer but containing 150 mM imidazole. The eluted fraction was stored at -20 °C until it was used. Treatment of RSO Membrane Vesicles with Thiol Reagents. Stock solutions of the thiol reagents N-ethylmaleimide (NEM) and 4-acetamido-40 -maleimidylstilbene-2,20 -disulfonic acid (AmdiS) were prepared freshly in water. The treatment of the reagents was stopped by addition of an equal concentration of dithiothreitol (DTT). The presence of DTT did not affect the initial rate of uptake in transport assays. RSO membranes at a concentration of 1 mg/mL were treated for the indicated times and at the indicated temperatures with the thiol reagents in 50 mM KPi (pH 7.0). Following treatment, RSO membranes were diluted twice into 50 mM KPi (pH 5.0) containing 140 mM NaCl. The pH of the resulting suspension was 6.0, and the suspension was immediately used for uptake measurements. Materials. NEM was purchased from Sigma-Aldrich BV (Zwijndrecht, The Netherlands), and AmdiS was purchased from Molecular Probes Europe BV (Leiden, The Netherlands). 14 14 L-[ C]Glutamate and [1,5- C]citrate were obtained from Amersham Pharmacia (Roosendaal, The Netherlands). RESULTS GGNG Sequence Motif in Region Vb of CitS. Each of the amino acid residues in sequence motif 184-GGNG-187 found in the Vb region of the Naþ-citrate transporter CitS of K. pneumoniae in the 2HCT family was substituted with a cysteine residue. The four mutants, G184C, G185C, N186C, and G187C, were tested for their ability to accumulate [1,5-14C]citrate in right-sideout (RSO) membrane vesicles prepared from E. coli DH5R cells expressing the mutants. Citrate uptake was assessed in the presence of a proton motive force (pmf) that was generated using the artificial potassium ascorbate/PMS electron donor system (Figure 2A). RSO membrane vesicles prepared from cells not expressing CitS lack citrate uptake activity because of the absence of a citrate transport system in the E. coli membrane. Membranes containing mutants G184C and G187C exhibited a similar uptake activity that was ∼10-15% of the activity of membranes containing wild-type CitS, while the G185C mutation resulted in complete lack of activity (Figure 2A). Mutation of the nonconserved position in the motif, N186 to Cys, resulted in an uptake activity of approximately 75% of that of wildtype CitS. Protein levels of the mutants in the membranes were estimated by small-scale purifications making use of the N-terminally fused His tag and Ni-NTA affinity chromatography (Figure 2B). All mutants exhibited similar expression levels as observed for wild-type CitS, indicating that the lower transport activity of the membranes containing the mutants was a consequence of the mutation in CitS rather than the lack of production or degradation of the proteins. It follows that the conserved Gly residues in the GGXG motif at positions 184 and 187 and especially Gly185 appear to be critical for the activity of the protein while mutation of the nonconserved N186 to Cys had an only marginal effect on the specific activity of CitS. N-Ethylmaleimide (NEM) is a small, membrane permeable thiol reagent. Treatment of the RSO membrane vesicles containing wild-type CitS, which contains five cysteine residues, with 1 mM NEM results in a slow inactivation of the protein to 10-20% of wild-type activity with a half-time of inactivation

Article

Biochemistry, Vol. 48, No. 31, 2009

7451

FIGURE 2: GGNG sequence motif in Vb of CitS of K. pneumoniae. (A) [1,5-14C]citrate uptake in RSO membrane vesicles containing CitS ([) and CitS mutants G184C (b), G185C (2), N186C (O), and G187C (0). (B) SDS-PAGE of partially purified CitS and the G184C, G185C, N186C, and G187C mutants purified from the RSO membranes used in the uptake assays shown in panel A. (C-H) Residual activity after treatment of RSO membranes containing CitS ([) and the mutants (9) G184C (C and F), N186C (D and G), and G187C (E and H) with 1 mM NEM (C-E) or 0.25 mM AmdiS (F-H) for the indicated time. Initial rates were expressed as the percentage of the initial rate catalyzed by untreated membranes.

of ∼4 min (23) (Figure 2C-E). In contrast, the three mutations in the GGxG motif that resulted in CitS proteins with residual activity (G184C, N186C, and G187C) rendered the proteins highly sensitive to NEM with inactivation half-times of